--datastore-conn-pool-read-max-open

did the trick and pods have stopped restarting/failing liveness checks. Thanks 🙂